Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20221006234138.1835739-1-keescook@chromium.org>
 <191ec24d-35d4-e4e5-85f7-d7301984e647@igalia.com> <202210171100.5BAC4A5CC8@keescook>
 <CAMj1kXHzrRTVcxb5+hgUPV3tjekPcDWzVf6cG_Mc9JJmYBz2Mw@mail.gmail.com>
 <202210171227.35ED875219@keescook> <CAMj1kXEJQ8gh-iXZNL8bNcmV=JCmKHNp5BnhYthhSOyR5h79_g@mail.gmail.com>
 <202210171237.DF5D4A3FD7@keescook> <CAMj1kXGmsJNg7En-55aRF+ApicPD_Opkh8Jw+oTorSOSO+cfuw@mail.gmail.com>
 <202210171307.32A5D9C07@keescook> <CAMj1kXHced1khwsrHqMUmECh_7irYOckFd+Sx3z9KSmsL7tPxw@mail.gmail.com>
 <202210171333.309A3D9@keescook> <CAMj1kXHTxc2PM1mz3pm-UEcmch9YG5QKF+JKAUNe9b+1L0OnfA@mail.gmail.com>
 <cb44c124-c940-6f5f-d195-bdf133008ba1@igalia.com>
In-Reply-To: <cb44c124-c940-6f5f-d195-bdf133008ba1@igalia.com>
From:   Ard Biesheuvel <ardb@kernel.org>
Date:   Mon, 17 Oct 2022 23:16:35 +0200
Message-ID: <CAMj1kXGA=kBdahDTuFDPzZKf5bbvxH2Ns7Ows+5D-PAx8rwP2w@mail.gmail.com>
Subject: Re: [PATCH] pstore: migrate to crypto acomp interface (take 2)
To:     "Guilherme G. Piccoli" <gpiccoli@igalia.com>
Cc:     Kees Cook <keescook@chromium.org>,
        Anton Vorontsov <anton@enomsg.org>,
        Colin Cross <ccross@android.com>,
        Tony Luck <tony.luck@intel.com>, linux-kernel@vger.kernel.org,
        linux-hardening@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, 17 Oct 2022 at 23:10, Guilherme G. Piccoli <gpiccoli@igalia.com> wrote:
>
> On 17/10/2022 18:01, Ard Biesheuvel wrote:
> > [...]
> >
> > In summary, we're better off sticking with the legacy comp interface,
> > but unfortunately, due to the way that has been plumbed into the
> > scomp/acomp with scatterlists version, that doesn't really help us get
> > rid of the memory overhead.
> >
>
> Out of curiosity, do you have a number here, like X kilobytes per active
> CPU?

2x128 KB per CPU, which makes 128  KB also the maximum input/output
size per request when using this interface. (SCOMP_SCRATCH_SIZE)

On my 32x4 CPU workstation, this amounts to 32 MB permanently locked
up for nothing when the pstore driver is loaded (with compression
enabled)